As negotiations advance among UN General Assembly members in New York, a revised draft resolution (rev1) has been issued outlining updated proposals for the terms of reference and modalities for the Scientific Panel on AI and Global Dialogue on AI Governance.
According to the revised draft, published on 15 May 2025, the Panel will comprise 40 members, to be appointed by an Independent Appointment Committee for a 3-year term. The Committee itself would consist of 10 members to be nominated by member states and elected by the Plenary on AI, and 10 members to be appointed by the Secretary-General.
The Panel’s task would be to issue timely and responsive scientific assessments, including one annual policy-relevant summary report synthesising and analysing existing research related to the opportunities, risks, and impacts of AI, and thematic briefs as it deems necessary. The Panel would also report on the progress of its work to the Plenary on AI twice annually.
The Plenary on AI is a new element introduced in the draft resolution (compared to the zero draft). It is intended to function as as a subsidiary of the General Assembly, and: meet twice annually; elect the 10 members of the Independent Appointment Committee nominated by Member States; discuss the work of the Panel and provide input to the Panel; and decide the date and location of the next Dialogue on AI Governance.
The Global Dialogue on AI Governance – involving governments and all relevant stakeholders – would have the mandate to: promote international cooperation on AI governance; identify ways on how AI can contribute to the implementation of the SDGs and closing the digital divide; discuss measures to respect, protect and promote human rights in the field of AI; address the social, economic, ethical, cultural, linguistic, and technical
dimensions of AI; share best practices, and lessons learned on AI governance; share information on existing AI capacity-building programs and identify capacity gaps; promote transparency, accountability and robust human oversight of AI in compliance with international law; strengthening safe, secure and trustworthy AI systems; and foster interoperability and compatibility of AI governance approaches.
To be convened annually, the Dialogue would include plenary meetings, with a high-level governmental segment, and multistakeholder thematic discussions. The first such dialogue is to take place in the margins of ITU´s AI for Good Summit in 2026. Each annual Dialogue would have two co-chairs appointed by the President of the General Assembly; they would be tasked with drafting a summary of each annual meeting, in consultation with member states and relevant stakeholders, and with informing the Panel of suggestions made by participants of the Dialogue relevant to their work. When the dialogue takes place in New York, the co-chairs would hold informal consultations in advance of the Dialogue on a declaration on AI governance to be adopted by the General Assembly.
The Panel and the Dialogue are to be reviewed during the High-Level Review of the Global Digital Compact at the eighty-second session of the General Assembly.
UN member states will discuss this version of the draft resolution on 20 May.